Cctv video abstraction and object detection for video surveillance system

Millions of closed-circuit television (CCTV) cameras are installed in streets and businesses throughout the world with the stated goal of reducing crime and increasing public safety. This leads to large collection of video data. For searching particular object or person from whole CCTV video data is too much time consuming. The proposed algorithm extracts key frames from these videos to overlook videos. From these key frames required frames are easily detectable and video can be played from that frame to analyze further in details. Extracting small number of key frames that can abstract the content of video is very important for efficient browsing and detection. The amount of data in video processing is significantly reduced by using video segmentation and key-frame extraction. This paper presents algorithm for key frames extraction with matching difference between two consecutive frames is computed with different weights. Shot boundaries are detected with automatic threshold. Key frame is extracted by using reference frame-based approach.